Paper data
Title:
Single autoterms selection for blind source separation in time-frequency plane Author(s): Holobar Ales, Faculty of electrical engineering and computer science, 2000 Maribor, Slovenia Févotte Cédric, IRCCyN, B.P 92101, 44321 cedex 03, France Doncarli Christian, IRCCyN, B.P 92101, 44321 cedex 03, France Zazula Damjan, Faculty of electrical engineering and computer science, 2000 Maribor, Slovenia Page numbers in the proceedings: Volume I pp 565-568 Session: Blind Source Separation / Independent Component Analysis (1/2)
Paper abstract
We address in this paper a method to perform blind separation of over-determined instantaneous linear mixtures of non-stationary sources. Our contribution is focused on the identification of single source autoterms which correspond to diagonal Spatial Time-Frequency Distribution of the sources vector with only one non-zero diagonal entry. These latter matrices are processed into an iterative joint-diagonalization scheme which provides an estimation of the mixing matrix.
Paper
|